InD v. East Berkshire Community Health NHS Trust [2005] UKHL 23, [2005] A.C. 373, Lord Nicholls considered the “radical suggestion” that the common law of negligence should jettison the concept of duty of care (the famous “fifth wheel on the coach”), making use instead of fault as the appropriate control mechanism for liability, analogous to the approach under the European Convention on Human Rights (ECHR). Despite recognising the attraction of this idea, Lord Nicholls expressed the reservation that to adopt it would lead to uncertainty in cases “where a person's acts or omissions do not render him liable in negligence for another's loss, even though this loss may be foreseeable.” A paradigm example of such a case is where A seeks to make B liable in negligence for C's deliberate wrongdoing, as in the Scottish appeal Mitchell v. Glasgow City Council [2009] UKHL 11, [2009] 2 W.L.R. 481.
Mitchell lived on a council estate, next door to the aggressive and anti-social Drummond, who had constantly harassed and threatened Mitchell after a row over loud music several years earlier. Their landlord, the council, knew of Drummond's behaviour. Eventually the council summoned Drummond to a meeting, where he was told that possession proceedings were being sought and that continued anti-social behaviour would result in his eviction. Drummond then lost his temper and became abusive, but later calmed down, apologised and left. Tragically, he went straight home and fatally assaulted Mitchell with an iron bar; he was later convicted of culpable homicide. Mitchell's widow and daughter sued the council in negligence, alleging that it owed a duty to warn Mitchell of the meeting with Drummond and that its failure to warn caused his death, because, if forewarned, he would have taken steps to avoid the attack. In addition, they alleged that the council was in breach of its positive obligation under Article 2 of the ECHR to safeguard Mitchell's right to life. The House of Lords concluded unanimously that neither claim was sustainable as a matter of law and both should be struck out.
All five Law Lords were clear that foreseeability of harm alone did not support a duty of care, especially where the defendant had merely failed to prevent deliberate harm by a third party. In this respect, they favoured Lord Goff's approach in Smith v. Littlewoods Ltd. [1987] 1 All E.R. 710, rather than Lord Mackay's test based on the degree of foreseeability of such acts. A positive duty would only arise where, exceptionally, this would be just, fair and reasonable, such as where the defendant created a source of danger (Haynes v. Harwood [1935] 1 K.B. 146), where the third party was under the control or supervision of the defendant (Dorset Yacht [1970] A.C. 1004) or where more generally the defendant had assumed responsibility to the victim to prevent the relevant sort of harm (Stansbie v. Troman [1948] 2 K.B. 48). No such exceptional features existed here; in contrast, policy reasons pointed away from liability. The duty could not logically be confined to social landlords but would cover social workers and private landlords too; it would be onerous and costly, would deter intervention to reduce such anti-social behaviour and lead to defensive practises. Lord Rodger also emphasised that a negligence duty would be inconsistent with the council's contractual obligations and the relevant statutory framework, neither of which required warning to be given to neighbours, even at the actual eviction stage which carries the greatest risk of retaliation by a disgruntled tenant.
The claim under Article 2 of the ECHR was equally doomed to failure. A public authority only comes under a positive duty to safeguard life if “it knew or ought to have known at the time of the existence of a real and immediate risk to the life of an identified individual”. But as Lord Hope pointed out, the facts did not come anywhere near satisfying this onerous test: Drummond had repeatedly threatened Mitchell in the past but had never used violence against him, and although he lost his temper at the meeting, he did not utter any threats against Mitchell nor was he armed with a weapon when he left. Interestingly, Lord Rodger rejected the ECHR claim for a different, more fundamental reason. He observed that the public body with the duty and the resources to prevent criminal violence was the local police force. The council might breach Article 2 if, for example, it allowed its housing stock to fall into such disrepair that tenants were at risk of life-threatening injuries, but not for failing to prevent criminal violence. So for Lord Rodger, even if council officials had been aware of a real and immediate risk to Mitchell's life (even if Drummond had left the meeting waving a gun and vowing to shoot Mitchell?), they would not have been under any Article 2 obligation to prevent it.
It is difficult to disagree with the result in Mitchell, despite considerable sympathy for the pursuers: Mitchell himself had urged the council to take steps to evict Drummond, whose behaviour was unpredictable, violent and irrational. There were good reasons not to warn in these circumstances, which might well outweigh the risk of not doing so. Overall, the gist seems to be that the council did not really do anything wrong – in other words, the council was not at fault. So, on reflection, maybe Lord Nicholls' radical suggestion of jettisoning the duty of care concept should be reconsidered.
It is clear that local housing authorities do owe a general duty of care to their tenants, so to make sense of Mitchell their Lordships had to contemplate duties of different types and analyse the detailed scope of those duties. But to ask whether there was a “duty” to take precisely the precaution the defendant is accused of failing to take, collapses the distinction between arguments of fault and of duty. It is sometimes said that the duty approach leads to certainty and clarity, especially for public authority defendants, but Mitchell is not crystal clear. For example, Lords Hope and Scott treated the case as involving a “pure omission”, whereas Lord Rodger and Baroness Hale expressly stated that it was not comparable with a classic pure omission case, because the council were “actors in this drama”.
The real culprit in the conceptual confusion is the over-used notion of foreseeability. Mitchell reminds us that foreseeability is not sufficient for the existence of a duty of care; likewise as a test for remoteness of damage it is essentially meaningless, given the breadth of the notion of the foreseeable type of harm and the fact that defendants are unlikely to be liable for foreseeable but deliberate interventions. Moreover, foreseeability is frequently treated as a synonym for fault in negligence, when it is no such thing (a sin arguably committed by Lord Nicholls when rejecting the “radical suggestion” in D v. East Berkshire). Of course the likelihood of harm, judged without the benefit of hindsight, is relevant to fault, but it is only one of a number of relevant factors to be weighed in the balance. This is why sometimes a defendant is at fault for failing to take a precaution against a risk of harm so infinitesimally small as to be, frankly, unforeseeable (as in The Wagon Mound (No. 2) [1967] 1 A.C. 617), whereas elsewhere a defendant is not at fault for running a clearly foreseeable risk of harm (as in Bolton v. Stone [1951] 1 All E.R. 1078). It is foreseeable that if trains run at 70 m.p.h. there will be more fatalities than if they are confined to 5 m.p.h., but it is nonsense to think that it is therefore per se blameworthy to run high speed trains – think of their social utility and the disastrous effect on the economy of confining them to walking pace. In fact, the test for fault is sophisticated, flexible and more than capable of accommodating the policy arguments against liability in cases like Mitchell, as well as ensuring, on other facts, that truly atrocious negligence results in liability.
